-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathAgent_7_with_defective_drone.py
218 lines (178 loc) · 9.98 KB
/
Agent_7_with_defective_drone.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
import random
import config
import utils
from prey import Prey
from predator import Predator
class Agent_7_wdd:
def __init__(self, prey_loc, predator_loc):
"""
Initializing the position of the Agent at locations where prey and predator are not present
Also initializes the belief state of the agent
Parameters:
self
prey_loc (int): Location of the prey
predator_loc (int): Location of the predator
"""
# Handling condition where prey and predator are spawned on the same location
list_to_choose_from = list(range(50))
if prey_loc == predator_loc:
list_to_choose_from.remove(prey_loc)
else:
list_to_choose_from.remove(prey_loc)
list_to_choose_from.remove(predator_loc)
self.curr_pos = random.choice(list_to_choose_from)
self.prev_pos = 999
# Initialize prey belief state
self.prey_belief_state = dict.fromkeys([i for i in range(50)], 1 / 49)
self.prey_belief_state[self.curr_pos] = 0
# print(f'Initial prey belief state: {self.prey_belief_state}')
# Initialize peadator belief state
self.predator_belief_state = dict.fromkeys([i for i in range(50)], 0)
self.predator_belief_state[predator_loc] = 1
# print(f'Initial predator belief state: {self.predator_belief_state}')
def move(self, arena, prey_loc, predator_loc):
"""
Moves Agent 1 according to the given priority
Parameters:
self
arena (dictionary): Adjacency list representing the graph
prey_loc (int): Location of prey
predator_loc (int): Location of Predator
"""
pos = utils.best_node(arena, self.curr_pos, prey_loc, predator_loc)
# Handling Sitting and praying case
if pos == 999:
pass
else:
self.prev_pos = self.curr_pos
self.curr_pos = pos
def begin(arena):
"""
Creates all the maze objects and plays number of games and collects data
Parameters:
arena (dict): Arena to use
Returns:
data_row (list): Results evaluated for the agent
"""
# Initiating game variables
game_count = 0
step_count = 0
# Initiating variables for analysis
win_count = 0
loss_count = 0
forced_termination = 0
data_row = []
number_of_games = config.NUMBER_OF_GAMES
forced_termination_threshold = config.FORCED_TERMINATION_THRESHOLD
prey_certainty = 0.0
predator_certainty = 0.0
while game_count < number_of_games:
# Creating objects
prey = Prey()
predator = Predator()
agent7_wdd = Agent_7_wdd(prey.curr_pos, predator.curr_pos)
step_count = 0
found_prey = False
found_predator = True
prey_certainty_counter = 0
predator_certainty_counter = 0
while 1:
print("In game Agent_7_wdd at game_count: ", game_count, " step_count: ", step_count)
print(agent7_wdd.curr_pos, prey.curr_pos, predator.curr_pos)
# Check if it knows where the predator is
if found_predator:
found_prey, node_surveyed = utils.survey_prey(agent7_wdd, prey)
if found_prey:
if random.random() <= 0.1:
found_prey = False
else:
found_predator, node_surveyed = utils.survey_predator(agent7_wdd, predator)
if found_predator:
if random.random() <= 0.1:
found_predator = False
# updating both belief states
agent7_wdd.prey_belief_state = utils.update_prey_belief_state(agent7_wdd.prey_belief_state, \
agent7_wdd.curr_pos, \
agent7_wdd.prev_pos, \
arena, \
found_prey, \
node_surveyed, \
'after_survey')
if max(agent7_wdd.prey_belief_state.values()) == 1:
prey_certainty_counter += 1
agent7_wdd.predator_belief_state = utils.update_predator_belief_state(agent7_wdd.predator_belief_state, \
agent7_wdd.curr_pos, \
agent7_wdd.prev_pos, \
arena, \
found_predator, \
node_surveyed, \
'after_survey')
if max(agent7_wdd.predator_belief_state.values()) == 1:
predator_certainty_counter += 1
believed_prey_curr_pos = utils.return_max_prey_belief(agent7_wdd.prey_belief_state, arena)
believed_predator_curr_pos = utils.return_max_predator_belief(agent7_wdd.predator_belief_state, arena)
# using the max belief node for prey
agent7_wdd.move(arena, believed_prey_curr_pos, believed_predator_curr_pos)
# Checking termination states
if agent7_wdd.curr_pos == prey.curr_pos:
win_count += 1
break
elif agent7_wdd.curr_pos == predator.curr_pos:
loss_count += 1
break
# update belief state
agent7_wdd.prey_belief_state = utils.update_prey_belief_state(agent7_wdd.prey_belief_state, \
agent7_wdd.curr_pos, \
agent7_wdd.prev_pos, \
arena, \
found_prey, \
node_surveyed, \
'after_agent_moves')
agent7_wdd.predator_belief_state = utils.update_predator_belief_state(agent7_wdd.predator_belief_state, \
agent7_wdd.curr_pos, \
agent7_wdd.prev_pos, \
arena, \
found_predator, \
node_surveyed, \
'after_agent_moves')
prey.move(arena)
agent7_wdd.prey_belief_state = utils.update_prey_belief_state(agent7_wdd.prey_belief_state, \
agent7_wdd.curr_pos, \
agent7_wdd.prev_pos, \
arena, \
found_prey, \
node_surveyed, \
'after_prey_moves')
# Checking termination states
if agent7_wdd.curr_pos == prey.curr_pos:
win_count += 1
break
predator.distracted_move(agent7_wdd.curr_pos, arena)
agent7_wdd.predator_belief_state = utils.update_predator_belief_state(agent7_wdd.predator_belief_state, \
agent7_wdd.curr_pos, \
agent7_wdd.prev_pos, \
arena, \
found_predator, \
node_surveyed, \
'after_predator_moves')
# Checking termination states
if agent7_wdd.curr_pos == predator.curr_pos:
loss_count += 1
break
step_count += 1
# Forcing termination
if step_count >= forced_termination_threshold:
forced_termination += 1
break
if step_count != 0:
prey_certainty += prey_certainty_counter / step_count
else:
prey_certainty = 0.0
if step_count != 0:
predator_certainty += predator_certainty_counter / step_count
else:
predator_certainty = 0.0
game_count += 1
data_row = ["Agent_7_wdd", win_count * 100 / number_of_games, loss_count * 100 / number_of_games,
forced_termination * 100 / number_of_games, prey_certainty * 100 / number_of_games, predator_certainty * 100 / number_of_games]
return data_row